检索结果-内蒙古大学图书馆

A Stacking Ensemble Learning Model for Mental State Recognition towards Implementation of Brain Computer Interface 6

A Stacking Ensemble Learning Model for Mental State Recognit...

引用

6th National-Foundation-for Science-and-Technology-Development (NAFOSTED) Conference on Information and Computer Science (NICS)

作者： Hoang-Anh The Nguyen Thanh Ha Le The Duy Bui Vietnam Acad Sci & Technol Inst Informat Technol Hanoi Vietnam Vietnam Natl Univ Univ Engn & Technol Hanoi Vietnam

ISBN: (数字)9781728151632

ISBN: (纸本)9781728151632

This paper presents a novel stacking ensemble learning model that aims at improving mental state classification for brain computer interface implementation. The proposed model combines machine learning based methods that use support vector machine, artificial neural network and deep learning with a model selection rule to classify EEG signals into accurate mental states. The proposed ensemble learning model is validated on an EEG dataset in which EEG signals are recorded from four subjects. Three mental tasks are turning mind into Zen condition (Neutral), imagining how to turn a light on (Light) and memorizing scientific paper content (Paper). Experimental results show that this ensemble learning model is robust and effective while comparing with other existing machine learning models and methods for the same purpose.

关键词： Ensemble learning EEG signals Brain computer interface Deep learning sparse autoencoder

来源：评论

学校读者我要写书评

暂无评论

Compressive Sensing and autoencoder Based Compressed Data Aggregation for Green IoT Networks

Compressive Sensing and Autoencoder Based Compressed Data Ag...

引用

IEEE Global Communications Conference (IEEE GLOBECOM)

作者： Zhang, Mingqiang Zhang, Haixia Yuan, Dongfeng Zhang, Minggao Shandong Prov Key Lab Wireless Commun Technol Jinan 250100 Peoples R China Shandong Univ Sch Control Sci & Engn Jinan 250100 Peoples R China

ISBN: (纸本)9781728109626

In cellular Internet of Things networks, massive access and highly dynamic traffic of machine type communication devices may cause network congestion, heavy energy consumption or even service unavailability. To reduce the amount of data transmission and improve the efficiency of data aggregation in sporadic machine type communications, we develop a two-stage compressed data aggregation scheme by jointly utilizing compressive sensing and sparse autoencoder. A deep compressive sensing network (DCSNet) is designed by using deep learning method to reconstruct signals from compressive measurements. Experimental results demonstrate that the proposed scheme can effectively reduce the data traffic, achieve higher reconstruction accuracy in different sparse conditions of signal and brake the obstacles that compressive sensing method has a large reconstruction error even when the measured data is small.

关键词： Data aggregation compressive sensing sparse autoencoder deep learning Internet of things

来源：评论

学校读者我要写书评

暂无评论

Intelligent condition monitoring method for bearing faults from highly compressed measurements using sparse over-complete features

引用

MECHANICAL SYSTEMS AND SIGNAL PROCESSING 2018年 99卷 459-477页

作者： Ahmed, H. O. A. Wong, M. L. D. Nandi, A. K. Brunel Univ London Dept Elect & Comp Engn Uxbridge UB8 3PH Middx England Heriot Watt Univ Malaysia Precinct 5 Putrajaya 62200 Malaysia Tongji Univ Coll Elect & Informat Engn Key Lab Embedded Syst & Serv Comp Shanghai Peoples R China

Condition classification of rolling element bearings in rotating machines is important to prevent the breakdown of industrial machinery. A considerable amount of literature has been published on bearing faults classification. These studies aim to determine automatically the current status of a roller element bearing. Of these studies, methods based on compressed sensing (CS) have received some attention recently due to their ability to allow one to sample below the Nyquist sampling rate. This technology has many possible uses in machine condition monitoring and has been investigated as a possible approach for fault detection and classification in the compressed domain, i.e., without reconstructing the original signal. However, previous CS based methods have been found to be too weak for highly compressed data. The present paper explores computationally, for the first time, the effects of sparse autoencoder based over-complete sparse representations on the classification performance of highly compressed measurements of bearing vibration signals. For this study, the CS method was used to produce highly compressed measurements of the original bearing dataset. Then, an effective deep neural network (DNN) with unsupervised feature learning algorithm based on sparse autoencoder is used for learning over-complete sparse representations of these compressed datasets. Finally, the fault classification is achieved using two stages, namely, pre-training classification based on stacked autoencoder and softmax regression layer form the deep net stage (the first stage), and re-training classification based on backpropagation (BP) algorithm forms the fine-tuning stage (the second stage). The experimental results show that the proposed method is able to achieve high levels of accuracy even with extremely compressed measurements compared with the existing techniques. (C) 2017 The Authors. Published by Elsevier Ltd.

关键词： Compressed sensing sparse over-complete representations Deep neural network sparse autoencoder Bearing fault classification Machine condition monitoring

来源：评论

学校读者我要写书评

暂无评论

HYPERSPECTRAL UNMIXING VIA WAVELET BASED autoencoder NETWORK 10

HYPERSPECTRAL UNMIXING VIA WAVELET BASED AUTOENCODER NETWORK

引用

10th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS)

作者： Yan, Bin Wu, Zebin Liu, Hongyi Xu, Yang Wei, Zhihui Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing 210094 Jiangsu Peoples R China

ISBN: (纸本)9781728152943

Hyperspectral unmixing is a hot topic in the field of remote sensing. Due to the limitation of spatial resolution and diversity of object distribution, hyperspectral image contains mixed pixels, which brings a great challenge for hyperspectral image processing. A novel wavelet based hyperspectral unmixing autoencoder network is proposed in this paper. In the framework of autoencoder network, multiscale wavelet coefficients of the signal are employed, which contribute to learn the intrinsic feature of endmember deeply. Moreover, the cost function of the network is designed according to the sparsity and nonnegative constraints of abundance, as well as the spectral fidelity. Experimental results on both simulated and real hyperspectral data sets demonstrate that the proposed method outperforms other state-of-the-art unmixing methods.

关键词： Hyperspectral unmixing sparse autoencoder wavelet domain

来源：评论

学校读者我要写书评

暂无评论

A Generative Model for sparse Hyperparameter Determination

引用

IEEE TRANSACTIONS ON BIG DATA 2018年第1期4卷 2-10页

作者： Wan, Zhiqiang He, Haibo Tang, Bo Univ Rhode Isl Dept Elect Comp & Biomed Engn Kingston RI 02881 USA

sparse autoencoder is an unsupervised feature extractor and has been widely used in the machine learning and data mining community. However, a sparse hyperparameter has to be determined to balance the trade-off between the reconstruction error and the sparsity of sparse autoencoder. Traditional sparse hyperparameter determination method is time-consuming, especially when the dataset is large. In this paper, we derive a generative model for sparse autoencoder. Based on this model, we derive a formulation to determine the sparse hyperparameter effectively and efficiently. The relationship between the sparse hyperparameter and the average activation of sparse autoencoder hidden units is also presented in this paper. Experimental results and comparative studies over numerous datasets demonstrate the effectiveness of our method to determine the sparse hyperparameter.

关键词： sparse autoencoder sparse hyperparameter feature extractor generative model

来源：评论

学校读者我要写书评

暂无评论

Feature extraction and physical interpretation of melt pressure during injection molding process

引用

JOURNAL OF MATERIALS PROCESSING TECHNOLOGY 2018年 261卷 50-60页

作者： Zhou, Xundao Zhang, Yun Mao, Ting Ruan, Yufei Gao, Huang Zhou, Huamin Huazhong Univ Sci & Technol State Key Lab Mat Proc & Die & Mold Technol 1037 Luoyu Rd Wuhan 430074 Hubei Peoples R China

Melt pressure is crucial in injection molding. A variety of pressure sensors have been installed in injection molding machines and molds to collect melt-pressure information. Many methods have been developed to extract the features of melt pressure. However, challenges exist for these feature extraction methods because they are, inevitably, extremely dependent on manual operation. In this study, an unsupervised feature extraction method using a sparse autoencoder is proposed to extract the features of melt pressure during injection molding. An injection molding model was applied to reinterpret the network structure of a sparse autoencoder to better understand the network structure. The feature curve, which is defined as the curve of weights between input and hidden units, was proved to be a significant indicator to measure how melt compression causes variations in injection pressure. Over 10,000 shots were conducted to verify the proposed feature extraction method. The experimental and simulation results show that the feature curve effectively extracts switching points (gate and velocity-pressure switchover), important indicators (viscosity index, holding pressure, and peak pressure) in the injection molding process and shows how the geometric features of a part affect the melt flow.

关键词： Feature extraction Feature physical interpretation sparse autoencoder Injection molding process

来源：评论

学校读者我要写书评

暂无评论

Deep learning of chroma representation for cover song identification in compression domain

引用

MULTIDIMENSIONAL SYSTEMS AND SIGNAL PROCESSING 2018年第3期29卷 887-902页

作者： Fang, Jiunn-Tsair Chang, Yu-Ruey Chang, Pao-Chi Ming Chuan Univ Dept Elect Engn 5 Deming Rd Taoyuan 33348 Taiwan Natl Cent Univ Dept Commun Engn 300 Jhongda Rd Taoyuan 32001 Taiwan

Methods for identifying a cover song typically involve comparing the similarity of chroma features between the query song and another song in the data set. However, considerable time is required for pairwise comparisons. In addition, to save disk space, most songs stored in the data set are in a compressed format. Therefore, to eliminate some decoding procedures, this study extracted music information directly from the modified discrete cosine transform coefficients of advanced audio coding and then mapped these coefficients to 12-dimensional chroma features. The chroma features were segmented to preserve the melodies. Each chroma feature segment was trained and learned by a sparse autoencoder, a deep learning architecture of artificial neural networks. The deep learning procedure was to transform chroma features into an intermediate representation for dimension reduction. Experimental results from a covers80 data set showed that the mean reciprocal rank increased to 0.5 and the matching time was reduced by over 94% compared with traditional approaches.

关键词： Cover song Music retrieval sparse autoencoder Descriptor Advanced audio coding

来源：评论

学校读者我要写书评

暂无评论

Classification of multiple power quality events via compressed deep learning

引用

INTERNATIONAL TRANSACTIONS ON ELECTRICAL ENERGY SYSTEMS 2019年第6期29卷 e12010.1-e12010.14页

作者： Liu, Hui Hussain, Fida Yue, Shen Yildirim, Ozal Yawar, Sheikh Junaid Jiangsu Univ Sch Elect & Informat Engn Zhenjiang Jiangsu Peoples R China Hamdard Univ Fac Engn Sci & Technol Karachi Pakistan Munzur Univ Dept Comp Engn Tunceli Turkey Sir Syed Univ Engn & Technol Dept Elect Engn Karachi Pakistan

This paper presents a recently established compressed sensing (CS) and sparse autoencoder (SAE) based on deep learning (DL) method for classification of single and multiple power quality disturbances (PQDs). The CS technique is paying considerable attention in recent years due to below sampling rate comparatively Nyquist sampling. Initially, the CS technique is applied to extract the features of PQD waveforms. The extracted features are applied as inputs to the sparse autoencoder based on DL for classification of nine single and 22 combined classes of PQDs. The DL helps to remove a redundant feature and improves classification performance. Finally, backpropagation is applied to fine-tune the entire network. The effectiveness of the proposed algorithm has been tested with more than 6580 numbers of real and synthetic single and multiple PQD data, and the results are recorded. High correct classification rate is obtained with noise and without noise level. Noise level was considered from 20 to 50 dB. The performance of the proposed technique has been assessed by comparing the results against recently reported methods. Results show that the proposed CS- and SAE-based DL algorithms can be efficiently used for single and multiple PQDs classifications.

关键词： compressed sensing deep machine learning power quality disturbances classification sparse autoencoder

来源：评论

学校读者我要写书评

暂无评论

Down image recognition based on deep convolutional neural network

引用

Information Processing in Agriculture 2018年第2期5卷 246-252页

作者： Wenzhu Yang Qing Liu Sile Wang Zhenchao Cui Xiangyang Chen Liping Chen Ningyu Zhang School of Cyber Security and Computer Hebei UniversityBaoding 071002PR China

Since of the scale and the various shapes of down in the image,it is difficult for traditional image recognition method to correctly recognize the type of down image and get the required recognition accuracy,even for the Traditional Convolutional Neural Network(TCNN).To deal with the above problems,a Deep Convolutional Neural Network(DCNN)for down image classification is constructed,and a new weight initialization method is ***,the salient regions of a down image were cut from the image using the visual saliency ***,these salient regions of the image were used to train a sparse autoencoder and get a collection of convolutional filters,which accord with the statistical characteristics of *** last,a DCNN with Inception module and its variants was *** improve the recognition accuracy,the depth of the network is *** experiment results indicate that the constructed DCNN increases the recognition accuracy by 2.7% compared to TCNN,when recognizing the down in the *** convergence rate of the proposed DCNN with the new weight initialization method is improved by 25.5% compared to TCNN.

关键词： Deep convolutional neural network Weight initialization sparse autoencoder Visual saliency model Image recognition

来源：评论

学校读者我要写书评

暂无评论

A Rotating Machinery Fault Diagnosis Method for High-speed Trains Based on Improved Deep Learning Network 7

A Rotating Machinery Fault Diagnosis Method for High-speed T...

引用

7th International Conference on Control Automation and Information Sciences (ICCAIS)

作者： Yang, Jing Xie, Guo Yang, Yanxi Li, Xin Xian Univ Technol Sch Automat & Informat Engn Xian 710048 Shaanxi Peoples R China Tianshui Normal Univ Sch Mechatron & Automot Engn Tianshui 741000 Peoples R China

ISBN: (纸本)9781538660201

Rotating machinery is an important part of highspeed train system. Any failure of such components will have a serious impact on the service itself and even jeopardize the safe and reliable operation of the train. However, the limited diagnostic accuracy has become the bottleneck of fault diagnosis. In this paper, an adaptive fault diagnosis method based on deep sparse autoencoder (SAE) network is proposed. Firstly, the original time domain signals collected by sensor are preprocessed and used as input to the diagnostic network. Secondly, the SAE fault diagnosis network model is built. Thirdly, the loss function of SAE is modified to improve the diagnostic accuracy and efficiency of network. Finally, experiments verify the feasibility and advantages of the proposed method.

关键词： sparse autoencoder ITS Non-stationary signal Fault diagnosis

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：